RE: [agi] Dog-Level Intelligence

2003-03-25 Thread Ben Goertzel

My feeling on dog-level intelligence is that the *cognition* aspects of
dog-level intelligence are really easy, but the perception and action
components are significantly difficult and subtle.

In other words, once a dog's brain has produced abstract patterns not tied
to particular environmental stimuli, the stuff it does with these patterns
is probably not all that fancy.  But the dog's brain is really good at
recognizing and enacting complex patterns, and doing this recognizing 
enacting in a coordinated way.

Peter Voss's (www.adaptiveai.com) approach to AI aims to emulate biological
evolution on Earth, in the sense that it wants to start with a dog-level
brain (very roughly speaking) and then incrementally build more cognition on
top of this.  This is a reasonable approach, to be sure.

But if I had to make a guess, I'd say this approach should probably begin
with robotics, with real sensors and actuators and a system embodied in a
real physical environment.  I am skeptical that simplistic simulated worlds
provide enough richness to support development of robust dog-level
intelligence... as perception and action oriented as dog intelligence is...

The current Novamente engineering plan is based on an opposite approach that
does not try to emulate biological evolution.   One *could* develop the
Novamente design in a biologically-inspired way, beginning with
perception/action and gradually adding cognition.  However, we are in fact
developing it in an opposite way, beginning with cognition in environments
requiring only very simple perception/action and then later adding more
advanced perception/action.  Of course, we realize that this
cognition-centric approach has been carried out in philosophically incorrect
ways by many AI researchers in the past, but, we believe we are not making
any of the basic errors that traditional cognition-focused AI researchers
have made.  Furthermore, I suspect that this is the only approach that has
any prayer of succeeding in the absence of either sophisticated robotic
sensors/actuators, or incredibly good VR environments...

-- Ben G






---
To unsubscribe, change your address, or temporarily deactivate your subscription, 
please go to http://v2.listbox.com/member/[EMAIL PROTECTED]


Re: [agi] Dog-Level Intelligence

2003-03-24 Thread Moshe Looks
Hi Margeret,

Margeret wrote:
 I don't mean to be rude here- but it may help you to follow some of
 the PSYCHE-D discussions on the Zombie argument. Their structure may
 help you to unravel some of your philosophical dead-ends. I don't mean
 to suggest that we don't all create them for our selves whentrying to
 study the mind and various types of cognitive systems. You may alos be
 interested in reading Daniel Dennett's argument on Ned Blocks' On
 being a Bat - you'll find that in Dennett's Consciousness Explained.
 (Yep- pretty arrogant title)

First of all, I'm not sure what gave you the idea that we were
discussing the consciousness of dogs, or lack thereof.  I am familiar
with the What's it like to be a Bat argument, but it seems to belong
in another discussion.  Did I ever claim that dogs were conscious, or
not conscious? Is there a specific place where I controversially
describe the subjective mental states of dogs?  And what philosophical
dead ends are you referring to?

 
 Personally I am of the opinion that it is always a disappointing
 scientific pursuit to speak about 'intelligence' in and of itself - or
 even compared with other species. Or to prove that various character
 or personality traits are indicative of its presence or absence. For
 example- saying that two agents have an IQ of 160 are equally
 intelligent is really a silly statement. And correlations are merely
 that- correlations.What we as scientists should be more interested in
 are causal dynamics. The structure and functions that interplay to
 allow for behaviour to appear intelligent. And in all of this we
 really are caught in evolution's stranglehold - because we only 'see'
 what works. I think it was either von Galsersfeld or Leo Apostel who
 claimed that The environment is held for extinction. What ever is
 not selected for is selected out. So, we have only our postulates.
 
Without agreeing or disagreeing with what you say here (and heaven
forbid, without being rude), I am not sure why you are telling me this. 
Are you trying to argue with something specific that someone else has
said in the discussion?  If so, what? To summarize the state of affairs,
I am not producing a dogie-IQ, rather a list of qualitative features. 
I think comparison between human and dog intelligence (for example) is
meaningful because we can point to qualitative features that humans and
dogs both possess, or that humans possess and dogs do not (and vice
versa).  All normal humans share the same qualitative features.  All
normal dogs share the same qualitative features. 

 In terms of the role of emotions in contributing towards reasoning,
 you may wish to do some serious neuropsyche reading. You could follow
 Le Douz, Damasio and that bunch just as a start. I don't mean this to
 be rude- it's just a very old and rather precarious Descartesian
 supposition that you refer to when you discuss the role of emotions
 and the 'extent' to which they contribute to reason. That's the same
 mad-making question as trying to figure out exactly how much of the
 visual cortex is involved when we do math, for e.g.
 
Let me quote exactly what I said: Its debatable to what extent emotions
actually contribute to
 intelligence ;-. Again, I'll have to think about this  I wrote
one sentence. I left the matter open for further discussion.  I used a
winking smiley, for crying out loud! I was not stating a serious
position!!!


 You allude to the need to integrate some theories of behaviour:
 Crikey Moshe! This stunned me! Haven't you heard of Psychology,
 Sociology, Political theory, history, theology etc etc... they all
 give Theories of Behaviour.
 
Did you read the original post, or the actual list I wrote
(http://www.republicofheaven.org/doglevel.html)? I am referring to
theories that dogs have of the way other entities behave.  Did you
understand this? If so, what stunned you?

 But perhaps I didn't get what you REALLY mean?

Maybe not, hope this clarifies things...

Cheers,
Moshe

---
To unsubscribe, change your address, or temporarily deactivate your subscription, 
please go to http://v2.listbox.com/member/[EMAIL PROTECTED]


Re: [agi] Dog-Level Intelligence

2003-03-21 Thread Brad Wyble


It might be easier to build a human intelligence than a dog intelligence simply 
because we don't have a dog's perspective and we can't ask them to reflect on it.  
Don't be quick to assume it would be easier just because they are less intelligent.  



-Brad

---
To unsubscribe, change your address, or temporarily deactivate your subscription, 
please go to http://v2.listbox.com/member/[EMAIL PROTECTED]